Method for identifying key element components in nuts and tracing origin

Abstract

The invention discloses a method for identifying key element components in nuts and tracing the origin of the nuts, which comprises the following steps of: processing content data of various elements of the nuts in one or more modes of multi-element correlation analysis, principal component analysis, partial least square discriminant analysis and orthogonal partial least square discriminant analysis; according to the method, feature distribution rules of elements of nuts from different producing areas in samples are obtained, a multi-element content analysis and producing area traceability identification model is constructed, the producing areas of the nuts are accurately identified, and the efficiency of identifying the producing area information of imported nuts is improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of traceability identification, and particularly relates to a method for identifying key element components in nuts and tracing origins. BACKGROUND

[0002] Nuts are the essence of plants, rich in nutrients, containing protein (12-36%), fat (≥40%), vitamins, trace elements, and functional ingredients such as phospholipids, polyphenols and flavonoids. Nuts are rich in unsaturated fatty acids (85-90% of total fat content), including linolenic acid, linoleic acid and other essential fatty acids for the human body, which are beneficial to improving human immunity. China imports a large number of nuts every year, and their origins are numerous, mainly concentrated in Australia, the United States, Africa and the Middle East. In the field of export customs clearance, the tax rates of nuts from different origins differ greatly, and it is easy to evade high customs duties through re-exports, that is, nuts from high-tariff countries are re-exported to countries with preferential tax rates without processing and value-added, in order to cheat preferential tax rates. Since the origin information cannot be identified by sensory means, this has caused great difficulty for port staff to identify nut origin information. Therefore, it is necessary to establish an effective method for identifying the origin of nuts. SUMMARY

[0003] To solve the above technical problems, the purpose of the present application is to provide a method for identifying key element components in nuts and tracing origins, which uses multi-element fingerprint to accurately identify the origin of nuts, so as to improve the efficiency of identifying the origin information of imported nuts.

[0004] To achieve the above-mentioned purpose of the application, the technical solutions adopted by the present application are as follows:

[0005] A method for identifying key element components in nuts and tracing origins, comprising

[0006] S1, standard sample preparation: selecting nut samples from different origins with clear regions as standard samples, and treating the standard samples by microwave digestion method;

[0007] S2, data acquisition of standard samples: detecting the nut standard samples by inductively coupled plasma emission spectrometry and inductively coupled plasma mass spectrometry respectively, and obtaining the multi-element content data of the nut standard samples;

[0008] S3, establishment of key component identification and origin traceability model: processing the multi-element content data of the nut standard samples by one or more of multi-element correlation analysis, principal component analysis, partial least squares discriminant analysis and orthogonal partial least squares discriminant analysis, determining the key origin discriminant elements corresponding to the nuts, and obtaining the element characteristic distribution rule of the nuts from different origins in the standard samples, to construct one or more identification models for multi-element content analysis of origin traceability;

[0009] S4, result prediction: after the nut sample to be detected is treated by the microwave digestion method, detection is performed by inductively coupled plasma optical emission spectrometry and inductively coupled plasma mass spectrometry, content data of multiple elements of the nut sample to be detected are obtained, and the content data are imported into a discrimination model of multi-element content analysis origin traceability to perform result prediction of origin traceability.

[0010] Preferably, in step S1, 0.500 g of the nut standard sample is weighed and placed in a microwave digestion tube, 4 mL of concentrated nitric acid is added, the microwave digestion tube is placed in a super microwave digestion instrument, super microwave digestion is performed by using a programmed temperature method, the sample after digestion is placed in a graphite digestion instrument to heat and drive off acid at 130 DEG C for 2 h, finally, the sample is transferred by using deionized water and the volume is adjusted to 25 mL.

[0011] Preferably, in step S2, the conditions of inductively coupled plasma optical emission spectrometry are as follows:

[0012] The power is 1300 W; the plasma gas flow rate is 12.0 L / min; the auxiliary gas flow rate is 0.20 L / min; the atomizing gas flow rate is 0.60 L / min; the peristaltic pump rotation speed is 50 r / min; the observation mode is axial and radial; and the carrier gas is 99.996% high-purity argon.

[0013] Preferably, in step S2, the conditions of inductively coupled plasma mass spectrometry are as follows:

[0014] The radio frequency power is 1550 W; the acquisition mode is peak jumping; the scan / reading is 20; the reading / repetition number is 1; the repetition number is 3; the collision pool is HE mode; the He flow rate is 4.5 mL / min; the atomizing gas flow rate is 0.90 L / min; the dilution gas flow rate is 0.20 L / min; the plasma gas flow rate is 1500 L / min; the ICP peristaltic pump rotation speed is 0.5 rps; the gas is 99.998% high-purity argon; and the sampling cone and the skimmer cone are platinum cones.

[0015] Preferably, in step S2, inductively coupled plasma optical emission spectrometry is used to detect 10 kinds of constant elements B, Mg, Al, Si, P, K, Ti, Mn, Zn and Rb in the nut sample; and inductively coupled plasma mass spectrometry is used to detect multiple elements including B, Mg, Al, Si, P, K, Ti, Mn, Zn and Rb in the nut sample.

[0016] Preferably, in step S3, multiple constant elements are selected, the correlation between the selected constant elements is analyzed by using two statistical methods of measuring linear correlation by using a Pearson correlation coefficient matrix and evaluating correlation by using a significance test, and it is determined that the mineral element content in nuts in different countries and regions has biological inherent regularity.

[0017] Preferably, in step S3, the multi-element content data is processed in the form of principal component analysis to construct a discrimination model for multi-element content analysis of origin traceability, which is used for rapid identification of the origin of nuts.

[0018] Preferably, in step S3, the multi-element content data is processed in the form of partial least squares discriminant analysis, and the discriminant variables with VIP>1.5 are selected as key elements to construct a discrimination model for multi-element content analysis of origin traceability, which is used for rapid identification of the origin of nuts.

[0019] Preferably, in step S3, the multi-element content data is processed in the form of partial least squares discriminant analysis, and the discriminant variables with VIP>1.5 are selected as key elements to construct a discrimination model for multi-element content analysis of origin traceability, which is used for rapid identification of the origin of nuts.

[0020] Preferably, the multi-element correlation analysis, principal component analysis, partial least squares discriminant analysis and orthogonal partial least squares discriminant analysis are combined to analyze the origin of nut samples from multiple angles and improve the accuracy.

[0021] In step S3, blind sample verification is also included, which comprises: selecting multiple nut verification samples from different regions, processing the nut verification samples by microwave digestion method and detecting the element content data by inductively coupled plasma atomic emission spectrometry and inductively coupled plasma mass spectrometry, introducing the obtained element content data into the discrimination model for multi-element content analysis of origin traceability in step S3, and verifying the origin traceability accuracy of the nut verification samples.

[0022] Beneficial effects:

[0023] The present application processes the multi-element content data of nuts by one or more of multi-element correlation analysis, principal component analysis, partial least squares discriminant analysis and orthogonal partial least squares discriminant analysis to obtain the element characteristic distribution rule of nuts from different origins in the sample, constructs a discrimination model for multi-element content analysis of origin traceability, accurately identifies the origin of nuts, and improves the efficiency of identifying the origin information of imported nuts. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 The principal component analysis score chart of pistachio nuts is shown;

[0025] Figure 2 The principal component analysis score chart of pistachio nuts is shown;

[0026] Figure 3 The partial least squares discriminant analysis score chart of pistachio nuts is shown;

[0027] Figure 4 The partial least squares discriminant analysis score chart of pistachio nuts is shown;

[0028] Figure 5 The score plot of orthogonal partial least squares discriminant analysis of pistachio is shown.

[0029] Figure 6 The score plot of orthogonal partial least squares discriminant analysis of pistachio is shown.

[0030] Figure 7 The origin traceability identification model of pistachio multi-element OPLS-DA of the United States and Iran is shown.

[0031] Figure 8 The origin traceability identification model of pistachio multi-element OPLS-DA of the United States and Iran is shown.

[0032] Figure 9 The PCA-X score plot of multi-country and multi-variety nuts is shown.

[0033] Figure 10 The PLS-DA score plot of multi-country and multi-variety nuts is shown.

[0034] Figure 11 The OPLS-DA score plot of multi-country and multi-variety nuts is shown. DETAILED DESCRIPTION

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, specific implementation manners of the present application will be described below with reference to the drawings. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained from these drawings without creative labor, and other embodiments can also be obtained.

[0036] The present application discloses a method for identifying key element components in nuts and tracing the origin, comprising

[0037] S1, standard sample preparation: select nut samples with different origins of specific regions as standard samples, and process the standard samples by microwave digestion method;

[0038] S2, data acquisition of standard sample: detect the nut standard sample by inductively coupled plasma emission spectrometry and inductively coupled plasma mass spectrometry respectively, and obtain the 58 element content data of the nut standard sample;

[0039] S3, key component identification and provenance traceability model establishment: the 58 element content data of the nut standard sample is processed by one or more of multielement correlation analysis, principal component analysis, partial least squares discriminant analysis, and orthogonal partial least squares discriminant analysis to obtain the element characteristic distribution rule of nuts of different origins in the standard sample, determine the key origin discriminant elements of nuts, and construct one or more multielement content analysis provenance traceability identification models;

[0040] S4, result prediction: the nut sample to be detected is treated by microwave digestion method, and then detected by inductively coupled plasma emission spectrometry and inductively coupled plasma mass spectrometry to obtain the 58 element content data of the nut sample to be detected, and then imported into the multielement content analysis provenance traceability identification model to predict the provenance traceability result.

[0041] In the present application, nuts include pistachios, almonds, cashews and macadamia nuts. It is easy to understand that other types of nuts can also be applicable based on the identification model constructed in the present application.

[0042] Preferably, in step S1, the nut standard sample is weighed and placed in a microwave digestion tube, concentrated nitric acid is added, and the tube is placed in a super microwave digestion instrument for super microwave digestion by a programmed temperature method. The digested sample is then placed in a graphite digestion instrument for heating and acid removal. Finally, the sample is transferred with deionized water and diluted to volume.

[0043] Preferably, in step S2, the conditions of inductively coupled plasma emission spectrometry are as follows:

[0044] The power is 1300W; the plasma gas flow rate is 12.0L / min; the auxiliary gas flow rate is 0.20L / min; the atomizing gas flow rate is 0.60L / min; the peristaltic pump speed is 50r / min; the observation mode is axial and radial; and the carrier gas is 99.996% high-purity argon.

[0045] Preferably, in step S2, the conditions of inductively coupled plasma mass spectrometry are as follows:

[0046] The radio frequency power is 1550W; the acquisition mode is peak jumping; the scan / read is 20; the read / repetition is 1; the repetition is 3; the collision pool is HE mode; the He flow rate is 4.5mL / min; the atomizing gas flow rate is 0.90L / min; the dilution gas flow rate is 0.20L / min; the plasma gas flow rate is 1500L / min; the ICP peristaltic pump speed is 0.5rps. The gas is 99.998% high-purity argon; and the sampling cone and skimmer cone are platinum cones.

[0047] Preferably, in step S2, 10 kinds of common elements, B, Mg, Al, Si, P, K, Ti, Mn, Zn, and Rb, are detected in the nut sample by inductively coupled plasma optical emission spectrometry; and 58 kinds of trace elements, Li, Be, B, Na, Mg, Al, Si, P, K, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, Ge, As, Se, Rb, Sr, Y, Zr, Nb, Mo, Pd, Ag, Cd, Sn, Sb, I, Cs, Ba, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, Hg, Tl, Pb, Bi, Th, and U, are detected in the nut sample by inductively coupled plasma mass spectrometry.

[0048] Preferably, in step S3, 5 kinds of common elements, such as Ca, Mg, P, Si, and K, are selected, and the correlation between the selected common elements is analyzed by two statistical methods, i.e., a linear correlation measured by a Pearson correlation coefficient matrix and a correlation evaluated by a significance test, to determine that the mineral element content in nuts in different countries and regions has biological intrinsic regularity.

[0049] Preferably, in step S3, the 58 kinds of element content data are processed in a principal component analysis manner to construct a discrimination model for multi-element content analysis of origin traceability, and K, Rb, B, and Mn are used as key origin discrimination indexes to distinguish almond produced in the United States and Australia.

[0050] Preferably, in step S3, the 58 kinds of element content data are processed in a partial least squares discriminant analysis manner, and a discriminant variable with a VIP value greater than 1.5 is selected as a key element to construct a discrimination model for multi-element content analysis of origin traceability, which is used for rapid identification of the origin of nuts.

[0051] For example, Na, K, Rb, and Mn are used as key elements for rapid identification of the origin of pistachio nuts.

[0052] For example, K, Rb, Ca, and Mg are used as key elements for rapid identification of the origin of almond nuts, in which K, Rb, and B are characteristic elements of almond nuts produced in the United States, and Ca, Mg, and Zn are characteristic elements of almond nuts produced in Australia.

[0053] Preferably, in step S3, the 58 kinds of element content data are processed in an orthogonal partial least squares discriminant analysis manner, discriminant variables with a VIP value in the top 15 are selected as key elements, and a discrimination model for multi-element content analysis of origin traceability is constructed, which is used for rapid identification of the origin of nuts.

[0054] For example, taking Na, Rb, K, B, Ca, Mn, Ba, Si, Al, Zn, Se, Cu, Mg, Fe, Ni as key elements, used for rapid identification of pistachio production areas, among them, Rb, K, B, Zn are characteristic elements of American pistachio, Ca, Si, Al, Mg are characteristic elements of Iranian pistachio;

[0055] For example, taking Rb, K, B, Na, Mn, Ba, Si, Al, Zn, Ca, Se, Cu, Mg, Fe, Ni as key elements, used for rapid identification of pistachio production areas, among them, Rb, K, B, Zn are characteristic elements of American pistachio, Ca, Si, Al, Mg are characteristic elements of Iranian pistachio;

[0056] Preferably, multi-element correlation analysis, principal component analysis, partial least squares discriminant analysis, and orthogonal partial least squares discriminant analysis are combined to analyze the origin of nut samples from multiple angles and improve accuracy.

[0057] In step S3, blind sample verification is also included, which comprises: selecting multiple nut verification samples from different regions, treating the nut verification samples by microwave digestion method and detecting the element content data by inductively coupled plasma emission spectrometry and inductively coupled plasma mass spectrometry, importing the obtained element content data into the identification model of multi-element content analysis of origin traceability in step S3, and verifying the origin traceability accuracy of the nut verification samples.

[0058] The technical solutions of the present application will be described in detail below with specific embodiments.

[0059] Real samples of soybean production areas in the United States, Australia, South Africa, Iran, etc. were collected by domestic main ports and overseas self-purchase. As of March 30, 2025, a total of 39 pistachio samples, 35 almond samples, 5 cashew samples and 9 macadamia nut samples were collected from relevant direct customs offices nationwide and abroad. The sample information is shown in Tables 1-4.

[0060] Table 1 Summary of pistachio sample information

[0061]

[0062]

[0063] Table 2 Summary of almond sample information

[0064]

[0065]

[0066] Table 3 Summary of macadamia nut sample information

[0067]

[0068]

[0069] Table 4 Summary of cashew sample information

[0070] No. Sample name Origin Source Y1 Cashew VNM Port Y2 Cashew VNM Port Y3 Cashew VNM Port Y4 Cashew VNM Port Y5 Cashew VNM Self-purchased

[0071] Multi-element analysis of nut samples: The contents of 58 elements such as Li, Be, B, Na, Mg, Al, Si, P, K, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, Ge, As, Se, Rb, Sr, Y, Zr, Nb, Mo, Pd, Ag, Cd, Sn, Sb, I, Cs, Ba, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, Hg, Tl, Pb, Bi, Th, U, etc. in nut samples were detected by inductively coupled plasma optical emission spectrometer (ICP-OES) and inductively coupled plasma-mass spectrometer (IPC-MS), combined with multivariate statistical analysis methods such as principal component analysis, partial least squares-discriminant analysis and orthogonal partial least squares regression analysis, to construct an ICP-MS provenance identification model for nuts. The details are as follows.

[0072] Equipment: Optima 8000 inductively coupled plasma optical emission spectrometer (PE, USA);

[0073] Agilent 8900 inductively coupled plasma mass spectrometer (ICP-MS) (Agilent Corporation);

[0074] MS 303TS / 02 electronic balance (Mettler-Toledo, USA);

[0075] Milstone super microwave digestion instrument (Milestone, Italy);

[0076] Graphite digestion instrument; Milli-Q ultrapure water system (Millipore Corporation);

[0077] The digestion tube for pretreatment before digestion is made of quartz.

[0078] Reagent material: concentrated nitric acid (UP grade, mass fraction 68%, Suzhou Crystal Company);

[0079] The experimental water is ultrapure water (resistivity > 18.2 MΩ·cm, 20℃);

[0080] 43 elements ICP-MS multi-element standard solution IV-ICP-MS-71A (10 mg / L, iNORGANic Ventures);

[0081] Multi-element standard solution IV-CCS-5 (100 mg / L, iNORGANic Ventures);

[0082] 3 elements ICP-MS rare earth mixed internal standard Re, Rh, In (100 μg / mL, inorganic Ventures);

[0083] Sodium, magnesium, phosphorus, potassium, calcium (1000 μg / mL) were purchased from Shanghai Anpu Experiment Technology Co., Ltd;

[0084] ICP analysis of GNM-M 292793-2013 mixed standard solution containing 29 elements (100 μg / mL, National Non-ferrous Metal and Electronic Material Analysis and Test Center). Tuning fluid containing Ba, Be, Ce, Co, Li, Mg, Rh, U (10 μg / mL, PerkinElmer Inc., USA);

[0085] Other reagents used were above analytical grade; Nitrogen: above 99.99%.

[0086] Sample pretreatment:

[0087] Microwave digestion method: according to GB 5009.268-2016 Foodstuff Multi-element Detection Method, accurately weigh about 0.500 g of sample (accurate value 0.001 g) into a microwave digestion tube, add 4 mL of concentrated nitric acid, and place the microwave digestion tube into a super microwave digestion instrument. Use the programmed heating method for super microwave digestion. The heating program is shown in Table 5. After digestion is completed, place it in a graphite digestion instrument to add acid chaser for about 2 h at 130°C. Finally, transfer the sample with deionized water and dilute to 25 mL. In addition, a blank test is performed.

[0088] Table 5 Super microwave heating program

[0089] Step Temperature / °C Time / min Pressure / KPa Power / W 1 180 4 15000 1500 2 210 8 15000 1500 3 210 10 15000 1500

[0090] Instrument analysis method:

[0091] Inductively coupled plasma optical emission spectrometry (ICP-OES): detection wavelength refers to GB 5009.268-2016, power is 1300 W; plasma gas flow is 12.0 L / min; auxiliary gas flow is 0.20 L / min; atomizing gas flow is 0.60 L / min; peristaltic pump speed is 50 r / min; observation mode is axial and radial; carrier gas is 99.996% high-purity argon.

[0092] Inductively coupled plasma mass spectrometry (ICP-MS): radio frequency power of 1550 W; acquisition mode of peak jumping; scan / read: 20; read / repetition: 1; repetition: 3; collision pool of HE mode; He flow rate of 4.5 mL / min; atomizing gas flow rate of 0.90 L / min; dilution gas flow of 0.20 L / min; isoplasma gas flow of 1500 L / min; ICP peristaltic pump speed of 0.5 rps; gas of argon of ≥99.998%; sampling cone and skimmer cone of platinum cone.

[0093] Data processing method:

[0094] ICP-OES and ICP-MS detection and analysis obtained 58 kinds of mineral element content data, through the descriptive statistical analysis method, induction and import SIMCA14.1 software (Switzerland Umetrics company), principal component analysis, partial least squares discriminant analysis, orthogonal partial least squares discriminant analysis, obtain the distribution rule of mineral element content in nuts of different producing areas, construct the identification model of nut multi-element content analysis producing area traceability.

[0095] Multi-element correlation analysis:

[0096] ICP-OES and ICP-MS were used to detect the content of mineral elements in soybean samples. Among them, ICP-OES was used to detect 10 kinds of constant elements such as B, Mg, Al, Si, P, K, Ti, Mn, Zn and Rb in nut samples, and ICP-MS was used to detect 58 kinds of trace elements such as Li, Be, B, Na, Mg, Al, Si, P, K, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, Ge, As, Se, Rb, Sr, Y, Zr, Nb, Mo, Pd, Ag, Cd, Sn, Sb, I, Cs, Ba, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, Hg, Tl, Pb, Bi, Th and U. Five kinds of constant elements such as Ca, Mg, P, Si and K were selected to explore the content correlation between each other in nut samples. Pearson correlation coefficient matrix was used to analyze the correlation between these elements, and the results are shown in Tables 6-7.

[0097] Table 6 Pearson correlation coefficient matrix correlation analysis of constant elements in pistachio samples

[0098] Element pair Correlation coefficient r p-value Significance Mg-Si -0.31 0.058 Marginal significant Mg-P 0.62 <0.001 Highly significant Mg-K 0.45 0.005 Significant Mg-Ca 0.08 0.642 Not significant Si-P -0.45 0.005 Significant Si-K -0.48 0.002 Significant Si-Ca -0.19 0.259 Not significant P-K 0.81 <0.001 Highly significant P-Ca 0.23 0.168 Not significant K-Ca 0.12 0.480 Not significant

[0099] Table 7 Pearson correlation coefficient matrix correlation analysis of constant elements in almond samples

[0100] Element pair Correlation coefficient r p-value Significance Mg-Si -0.21 0.229 Not significant Mg-P 0.18 0.312 Not significant Mg-K 0.42 0.013 Significant Mg-Ca 0.15 0.403 Not significant Si-P -0.14 0.433 Not significant Si-K -0.37 0.03 Significant Si-Ca -0.4 0.018 Significant P-K 0.46 0.006 Significant P-Ca -0.07 0.699 Not significant K-Ca -0.17 0.344 Not significant

[0101] According to the results of Table 6-7, P and K elements in pistachio samples showed a very strong positive correlation, Mg and P elements had a moderate positive correlation. Si and K, Si and P showed significant negative correlation; Ca was generally weakly associated with other elements and was not significant. In almond samples, P and K, Mg and K showed a moderate positive correlation, while Si and K, Si and Ca showed significant negative correlation.

[0102] It can be seen that for pistachio and almond, P and K are positively correlated (pistachio correlation coefficient r = 0.81, almond correlation coefficient r = 0.46), indicating the synergistic effect of these two elements in plant physiological processes, which may be related to energy metabolism and ion balance. The negative correlation of Si and K, Ca may indicate that silicon may compete with these cations in the process of cell wall deposition or reflect the influence of absorption and transport mechanisms in different planting regions. Based on the data in Table 6-7, it can be reasonably speculated that the mineral element content in nuts in different countries and regions has biological internal regularity, proving that the use of element content for provenance analysis is scientific and reasonable.

[0103] Construction of provenance model:

[0104] Principal component analysis model (PCA analysis model):

[0105] From Table 1, 58 kinds of element content data of 38 pistachio samples (29 from the United States, 8 from Iran, and 1 from Australia) were selected as grouping variables, and the element content data were standardized by Z-score to eliminate the influence of dimension. The PCA analysis results are shown in Table 8, and the first two principal components cumulatively explained 51.1% of the total variance.

[0106] Table 8 Principal component analysis of pistachio

[0107]

[0108]

[0109] In Table 8, the main contribution elements of PC1 were Na (0.42), K (0.39), and P (0.37) with positive load, and Ca (-0.31) and Zn (-0.29) with negative load. The main contribution elements of PC2 were Al (0.35) and Ti (0.33) with positive load, and Rb (-0.41) and 133Cs (-0.38) with negative load. The PCA score plot is shown in Figure 1 .

[0110] The pistachio samples from the United States and Iran showed a good separation trend on the PC1 axis, i.e., the horizontal coordinate axis, with a small amount of samples (about 15%) overlapping in the boundary region of the two categories. The samples from the United States were more widely distributed on PC1, showing greater variability in elemental content. The samples from Iran were relatively concentrated on the negative half of PC2 (PC2 axis is the vertical coordinate axis). Through load analysis, the most important elements for distinguishing the production area were identified as follows: Na, the samples from the United States were generally significantly higher than those from Iran (p<0.001); K, the samples from Iran were slightly higher but the difference was not significant (p=0.056); Zn, the samples from Iran were significantly higher than those from the United States (p<0.01); Rb, the samples from the United States were significantly higher than those from Iran (p<0.001); Al, the samples from Iran were significantly higher than those from the United States (p<0.05). Evaluation of the model found that the cross-validation accuracy was 85.3%, the correct classification rate of the United States samples was 89.7%, and the correct classification rate of Iran was 77.8%. The difference in sodium content may be due to the seasoning processing of some samples, the difference in aluminum content may be related to the mineral composition of the soil, and the difference in potassium and phosphorus content may reflect different fertilization strategies. Through principal component analysis load, the key elements for distinguishing the production area of pistachio are Na, K, Zn, Rb, and Al.

[0111] The data of 34 almond samples selected from Table 2 were processed, and 58 element detection results were used for modeling analysis, and the principal component analysis results are shown in Table 9.

[0112] Table 9 Principal component analysis of almond

[0113] Principal component Proportion of explained variance Cumulative proportion of explained variance PC1 32.5％ 32.5％ PC2 15.8％ 48.3％ PC3 9.2％ 57.5％ PC4 7.1％ 64.6％ PC5 5.3％ 69.9％

[0114] The first five principal components explained about 70% of the total variance, and the principal component load analysis results showed that PC1 was mainly affected by the following elements: K, Rb, Na, and B with positive load; and Ca, Mg, and Zn with negative load. PC2 was mainly affected by the following elements: Mn, Al, and Fe with positive load; and Cu, Zn, and Ba with negative load. By observing the score plot of the first two principal components, i.e. Figure 2 It can be observed that the almond samples from the United States and Australia have a certain degree of separation trend in the principal component space. The samples from the United States generally have higher scores on PC1, which is related to the higher content of elements such as K, Ba, Na, and B. The samples from Australia have higher scores on PC2, which is related to the content of elements such as Mn, Al, and Fe. There is a certain overlap region in the figure, indicating that the almonds from the two production areas have similarities in elemental composition.

[0115] The key difference elements were analyzed, and the following elements played an important role in the origin differentiation. K and Rb: the content of the American sample was significantly higher than that of the Australian sample; B: the content of the American sample was higher; Mn and Al: the content of part of the Australian sample was higher; Ca and Mg: the content of the Australian sample was relatively high. The principal component analysis based on multi-element analysis can partially distinguish the pistachio produced in the United States and Australia, and K, Rb, B, Mn and other elements are the key indicators for distinguishing the origin.

[0116] The partial least squares discriminant analysis model (PLS-DA analysis model) is as follows:

[0117] The pistachio data was preprocessed, and 58 effective element variables were analyzed. The origin information (Iran / United States / Australia) was converted into a binary variable (0 / 1). The PLS-DA analysis results are as follows:

[0118] The model performance evaluation parameters are as follows: X variable explained variance (R2X): 0.42; Y variable explained variance (R2Y): 0.89; prediction ability (Q2): 0.85.

[0119] The cross-validation results show that the model correct classification rate is 90.3%; the sensitivity (identifying Iranian samples) is 91.1%; and the specificity (identifying American samples) is 90.6%.

[0120] Variable importance analysis (VIP): the top 10 discriminant variables (VIP>1.5) are as follows, namely Na, K, Rb, B, Mg, Al, Mn, Ca, Fe and Cu.

[0121] The identification model score diagram is as shown in Figure 3 The score diagram of the first two latent variables can be observed: the pistachio samples from Iran and the United States are obviously separated in the PLS-DA space. The American samples have a higher score on the first latent variable (LV1), which is mainly related to the content of Na, K, Rb and other elements. The Iranian samples have a higher score on the second latent variable (LV2), which is related to the content of Mn, Al, Fe and other elements. Only a small number of samples are located in the overlapping area, indicating that the model established in the present application has good discrimination ability. Through analysis of the VIP value and the loading diagram, the following elements play an important role in the origin differentiation.

[0122] The elements with higher content in American samples were: Na content was significantly higher than that in Iranian samples; K content was higher on average; Rb content was significantly higher than that in Iranian samples; B content was higher. The elements with higher content in Iranian samples were: Mn content was significantly higher in some samples; Al content was higher on average; Fe content was higher; Ca content was higher in some samples. The results of PLS-DA model were explained as follows: PLS-DA model showed good discrimination ability (Q2=0.85), and the cross-validation accuracy was as high as 92.3%, indicating that multi-element fingerprint could effectively distinguish pistachio from Iran and the United States. The characteristics of American pistachio were higher content of alkali metal elements such as Na, K, and Rb, which may be related to the soil composition or fertilization method in some planting areas in the United States. The characteristics of Iranian pistachio were higher content of transition metal elements (Mn, Fe, etc.) and Al, which may reflect the soil mineral composition characteristics of the planting area in Iran. A small number of American samples (K9-K12) had abnormally high Na content, which may be due to the addition of sodium-containing seasonings during processing. Some Iranian samples (K3, K5) had significantly higher Mn content than other samples, which may come from specific planting areas. Model validation and stability were tested by permutation test, and 200 times of permutation test showed p<0.001, which excluded the possibility of overfitting. All variables with VIP>1 passed the significance test (p<0.05).

[0123] The PLS-DA model based on key elements such as Na, K, Rb, and Mn can be used for rapid identification of pistachio origin. This method can be used as a supplement to traditional traceability methods to improve the objectivity and accuracy of origin identification.

[0124] Multi-element fingerprint combined with PLS-DA analysis can effectively distinguish pistachio from Iran and the United States, with a correct rate of over 90%. This method can be applied to pistachio quality control and origin authenticity verification, providing a scientific basis for market supervision.

[0125] The multi-element data of pistachio were preprocessed, and 58 effective element variables were analyzed. Z-score standardization (mean=0, standard deviation=1) was used. The PLS-DA model score plot is shown in FIG. 3. Figure 4The model performance indicators R2X (X variable explained variance) is 0.38, R2Y (Y variable explained variance) is 0.91. Q2 (predictive ability) is 0.87, cross-validation accuracy is 93.9%, sensitivity (identify the United States sample) is 95.2%, specificity (identify the Australian sample) is 92.3%. LV1 explained variance is 58.7%, LV2 explained variance is 21.3%. Variable importance projection (VIP) analysis results show that: VIP>1.5 key discriminant elements are as follows: K, Rb, B, Na, Ca, Mg, Al, Zn, Cu. Score plot analysis on the United States sample set in LV1 positive half axis (positive correlation with K, Rb, B), Australian sample set in LV1 negative half axis (positive correlation with Ca, Mg).

[0126] A9 sample (United States) deviates from the main group on LV2, showing abnormally high Rb content. A16 sample (Australia) shows abnormally low Al content. The United States almond has significantly high elements of K, Rb and B, which may be due to the fact that the soil in California is rich in potassium elements, the difference in local agricultural fertilization practice and the influence of climate conditions on element absorption. The Australian almond has significantly high elements of Ca, Mg and Zn, which may be due to the characteristics of soil mineral composition in Australia, element accumulation under arid climate conditions, and element absorption characteristics caused by variety differences. The above results all show that there are regional differences in almond elements.

[0127] Model validation and stability are tested by replacement test. 200 replacement tests show p<0.001, which excludes the possibility of overfitting. All variables with VIP>1 pass the significance test (p<0.05).

[0128] The above analysis results show that monitoring key elements such as K, Rb, Ca, Mg, etc. can be used as origin identifiers. Threshold criteria are established: such as Rb>10mg / kg can be initially determined as American production.

[0129] PLS-DA model can effectively distinguish American and Australian almonds, with an accuracy of 93.9%. K, Rb, B are characteristic elements of American almond; Ca, Mg, Zn are characteristic elements of Australian almond. This analysis method can be used as a reliable technical means for identifying the authenticity of almond origin, and provides a scientific basis for traceability and quality control.

[0130] Orthogonal partial least squares discriminant analysis model (OPLS-DA analysis model):

[0131] The collected pistachio multi-element detection data was subjected to orthogonal partial least squares discriminant analysis (OPLS-DA) to distinguish pistachio samples from different origins. The data contains the content of 58 elements of 39 samples (8 Iranian samples, 30 American samples and 1 Australian sample). The score plot is shown inFigure 5 The model performance indicators are R2X=0.723 (cumulative explained variance of X variables), R2Y=0.891 (cumulative explained variance of Y variables), and Q2=0.843 (model prediction ability). The results of the permutation test show that p<0.001 for 200 permutations, which excludes the possibility of overfitting. The top 15 key elements in terms of variable importance projection (VIP) values are as follows: Na, Rb, K, B, Ca, Mn, Ba, Si, Al, Zn, Se, Cu, Mg, Fe, and Ni. The loading plot shows that the first principal component t[1] explains 65.7% of the X variable variation and 83.7% of the Y variable variation. The positively correlated elements of the American samples are Na, Rb, K, and B, while the negatively correlated elements of the Iranian samples are Ca, Ba, Si, and Al. The distribution characteristics of the key elements show that Na and Rb in the upper right quadrant are strongly positively correlated with the American samples, Ca and Si in the upper left quadrant are correlated with the Iranian samples, and the elements in the central region, such as Hg, have no significant contribution to classification.

[0132] The characteristic elements for origin discrimination are as follows: The characteristic elements of American pistachio nuts are significantly enriched in Na, Rb, and K. Since Na may be added through processing, the typical element combination is Rb-K-B-Zn. The characteristic of Iranian pistachio nuts is high Ca, Si, and Al content. The characteristic element combination is Ca-Si-Al-Mg, which may be related to the local limestone parent soil.

[0133] The model validation accuracy results show that the training set is 96.2%, and the cross-validation is 90.5%. The OPLS-DA model successfully establishes a reliable origin discrimination system through the fingerprint of 58 elements, and the key difference elements reflect the significant influence of different geographical environments on the absorption of mineral elements. American pistachio nuts show a typical rubidium enrichment characteristic, while Iranian samples show a calcium and silicon dominant pattern, providing an effective chemical fingerprint method for pistachio geographical tracing. The model has excellent prediction ability (Q2>0.8) and can be used for origin tracing verification.

[0134] The collected pistachio multi-element detection data were subjected to orthogonal partial least squares discriminant analysis (OPLS-DA) to distinguish pistachio samples from different origins. The data included the contents of 58 elements in 33 samples (20 American samples and 13 Australian samples). The score plot is shown in Figure 6The model performance indicators are R2X=0.682 (cumulative explanatory rate of X variables), R2Y=0.873 (cumulative explanatory rate of Y variables), and Q2=0.812 (model prediction ability). The results of the permutation test show that p<0.001 in 200 permutation tests, which excludes the possibility of overfitting. The top 15 key elements in terms of VIP values are Rb, K, B, Na, Mn, Ba, Si, Al, Zn, Ca, Se, Cu, Mg, Fe, and Ni. The loading plot shows that the first principal component t[1] explains 62.3% of the X variable variation and 81.7% of the Y variable variation. The positively correlated elements of the American samples are Rb, K, B, and Na. The negatively correlated elements of the Australian samples are Mn, Ba, and Si. The distribution characteristics of the key elements are as follows: Rb and K are strongly positively correlated with American samples in the upper right quadrant. Mn and Ca are characteristic of Australian samples in the upper left quadrant. Elements in the central region, such as Hg, do not significantly contribute to classification.

[0135] In summary, the characteristic elements of American almonds are significantly enriched in Rb, K, B, Na, and Zn. This may reflect the potassium-rich soil environment and specific fertilization strategies. The characteristics of Australian almonds are high in Mn and Ca. The characteristic element combination is Mn-Ca-Si-Al. This may be related to the local basalt parent material soil. The traceability marker can be selected as Rb content (USA>10mg / kg, AUS<510mg / kg). The auxiliary marker element is the Mn / K ratio (AUS>0.005, USA<0.003). The verification accuracy results show that the training set is 94.7%, and the cross-validation accuracy is 89.3%. Sample A16 deviates from the main cluster due to abnormally low Al content.

[0136] The OPLS-DA model successfully establishes a reliable origin discrimination system through the fingerprint of 58 elements. The key difference elements reflect the long-term impact of different geographical environments on the absorption of mineral elements, providing an effective chemometric method for almond geographical traceability.

[0137] Optimization of prediction models for the two countries and blind sample verification:

[0138] From the comparison of the above three modeling analysis results, it is found that the orthogonal partial least squares model has good discrimination for nuts from the two countries. According to the analysis results, variables with VIP greater than 1 are selected for modeling, and Na element is excluded, as well as suspicious samples. The final OPLS-DA model is shown in Figure 7 and Figure 8The two-country model can significantly distinguish pistachio samples from the United States and Iran. It can also significantly distinguish almond samples from the United States and Australia. In order to further verify the determination accuracy of the nut origin traceability identification model, the experiment selected blind samples (4 Iranian pistachios, 10 American pistachios, 10 Australian almonds, and 10 American almonds) for verification, and the verification results showed that the identification accuracy of American pistachios was 90.0%, the identification accuracy of Iranian pistachios was 100%, the identification accuracy of American almonds was 90.0%, and the identification accuracy of Australian almonds was 90.0%.

[0139] Multi-country and multi-variety nut prediction model and blind sample verification:

[0140] Multivariate statistical analysis was performed on the multi-element determination results of pistachios, almonds, cashews, and macadamia nuts from four major countries, and 21 elements with greater contribution to the differentiation of the multi-country and multi-variety nut origin traceability model were selected, including Li, B, Mg, K, Ca, Mn, Fe, Ni, Cu, Zn, Ga, Se, Rb, Sr, Cd, Cs, Ba, Eu, Hg, Bi, and U. The above 21 elements were used to establish PCA-X, PLS-DA, and OPLS-DA models, and the score plots are shown in Figure 9-11 Comparing the three score plots, in the PCA-X and PLS-DA score plots, American almonds and Australian almonds are difficult to distinguish. In the OPLS-DA score plot, the four nuts can be distinguished according to the country of origin.

[0141] Further analysis of the OPLS-DA model showed that R2X was 0.880, R2Y was 0.909, and Q2 was 0.879. This indicates that the model has strong data interpretation and prediction capabilities (Q2>0.5), and is suitable for origin identification. The permutation test is as follows: randomly permute the group labels 200 times to generate the R2Y and Q2 distributions of the random model. The results show that the R2Y and Q2 of the original model are significantly higher than those of the random model (p<0.01), indicating that the model is not over-fitted and the results are reliable.

[0142] Analysis of the key elements related to the origin showed that Rb was significantly higher in American samples (mean 16.5 vs IRAN 4.2), Mg was extremely high in Vietnamese samples (mean 2800 vs <1300 in other samples), Australian samples contained high Ba (mean 3.2 vs 0.5 in Vietnamese samples), American samples contained high Zn (mean 22.5 vs 18.1 in Iranian samples), and Australian samples contained extremely low Pb (mean 0.0003 vs 0.001 in other samples).

[0143] The model is verified, and the classification accuracy is 93% for the training set and 88% for the test set. Cross-validation: Q2>0.8, indicating that the model is robust for new sample prediction. In order to further verify the judgment accuracy of the nut origin traceability identification model, the experiment selects blind samples (4 Iranian pistachios, 5 American pistachios, 8 Australian almonds, 5 American almonds, 2 Australian macadamia nuts, and 2 Vietnamese cashews) for verification. The verification results show that the identification accuracy of American nuts is 90.0%, the identification accuracy of Iranian nuts is 100%, the identification accuracy of Australian nuts is 90.0%, and the identification accuracy of Vietnamese nuts is 100%.

[0144] The present application analyzes the content of 58 mineral elements in the result samples of different countries and regions, and constructs a PCA, PLS-DA and OPLS-DA origin traceability identification model for nut multi-element content analysis according to the content of mineral elements. It is found that the OPLS-DA origin traceability identification model for multi-element content analysis can significantly distinguish pistachio and almond nut samples from different countries. The experiment selects blind samples for verification, and the verification results show that the identification accuracy of American pistachios is 90.0%, the identification accuracy of Iranian pistachios is 100%, the identification accuracy of American almonds is 90.0%, and the identification accuracy of Australian almonds is 90.0%. At the same time, OPLS-DA analysis of 4 kinds of nut samples from different countries can also distinguish nuts from different origins, and the origin traceability accuracy is more than 90%. In summary, the nut multi-element analysis origin traceability identification model of the present application can accurately determine the origin of unknown samples.

[0145] The above provides a detailed description of the embodiments of the present application. The principles and implementation methods of the present application are described by applying specific examples. The above description of the embodiments is only used to help understand the core idea of the present application. It should be noted that for ordinary skilled persons in the technical field, without departing from the principles of the present application, the present application can be improved and modified, and these improvements and modifications also fall within the protection scope of the claims of the present application.

Claims

1. A method for identifying key elemental components and tracing the origin of nuts, characterized in that, include S1. Preparation of standard samples: Select nut samples from different production areas with clear regions as standard samples, and process the standard samples by microwave digestion method; S2. Data acquisition of standard samples: Nut standard samples were detected by inductively coupled plasma atomic emission spectrometry and inductively coupled plasma mass spectrometry to obtain the content data of multiple elements in the nut standard samples; S3. Key component identification and origin traceability model establishment: The element content data of nut standard samples are processed by one or more methods such as multi-element correlation analysis, principal component analysis, partial least squares discriminant analysis, and orthogonal partial least squares discriminant analysis to obtain the distribution pattern of element characteristics of nuts from different origins in the standard samples, and to construct one or more multi-element content analysis origin traceability identification models. S4. Result Prediction: After the nut samples to be tested are processed by microwave digestion, the elemental content data of the nut samples are obtained by inductively coupled plasma atomic emission spectrometry and inductively coupled plasma mass spectrometry. The data are then imported into the multi-element content analysis and origin traceability identification model to predict the origin traceability results.

2. The method for identifying key elemental components and tracing origin in nuts according to claim 1, characterized in that, In step S1, weigh the nut standard sample and place it in a microwave digestion tube, add concentrated nitric acid, place it in a super microwave digester, and perform super microwave digestion using a programmed temperature rise method. After digestion, place the sample in a graphite digester to heat and remove the acid, and finally transfer the sample with deionized water and make up the volume.

3. The method for identifying key elemental components and tracing origin in nuts according to claim 1, characterized in that, In step S2, the conditions for inductively coupled plasma atomic emission spectrometry are: The power is 1300W; the plasma gas flow rate is 12.0L / min; the auxiliary gas flow rate is 0.20L / min; the atomizing gas flow rate is 0.60L / min; the peristaltic pump speed is 50r / min; the observation methods are axial and radial; the carrier gas is 99.996% high-purity argon.

4. The method for identifying key elemental components and tracing origin in nuts according to claim 1, characterized in that, In step S2, the conditions for inductively coupled plasma mass spectrometry are: RF power: 1550W; Acquisition mode: peak skipping; Scan / reading: 20; Reading / repetition count: 1; Repetition count: 3; Collision cell: HE mode; He flow rate: 4.5 mL / min; Nebulizer gas flow rate: 0.90 L / min; Dilution gas flow rate: 0.20 L / min; Plasma gas flow rate: 1500 L / min; ICP peristaltic pump speed: 0.5 rpm. Gas: 99.998% high-purity argon; Sampling cone and retrieval cone: platinum cone.

5. The method for identifying key elemental components and tracing origin in nuts according to any one of claims 1-4, characterized in that, In step S2, 10 major elements (B, Mg, Al, Si, P, K, Ti, Mn, Zn, and Rb) in the nut sample are detected by inductively coupled plasma atomic emission spectrometry (ICP-AES); and multiple elements including B, Mg, Al, Si, P, K, Ti, Mn, Zn, and Rb in the nut sample are detected by inductively coupled plasma mass spectrometry (ICP-MS).

6. The method for identifying key elemental components and tracing origin in nuts according to claim 1, characterized in that, In step S3, multiple constant elements are selected, and the correlation between the selected constant elements is analyzed using two statistical methods: the Pearson correlation coefficient matrix is ​​used to measure linear correlation, and the correlation is evaluated through significance tests. This determines that the mineral element content in nuts from different countries and regions has an inherent biological regularity.

7. The method for identifying key elemental components and tracing origin in nuts according to claim 1, characterized in that, In step S3, the content data of multiple elements are processed by principal component analysis to construct an identification model for origin traceability based on multi-element content analysis, which is used for rapid identification of nut origin.

8. The method for identifying key elemental components and tracing origin in nuts according to claim 1, characterized in that, In step S3, the data on the content of multiple elements are processed by partial least squares discriminant analysis. Discriminant variables with VIP > 1.5 are selected as key elements to construct an identification model for the origin traceability of multi-element content analysis, which is used for the rapid identification of the origin of nuts.

9. The method for identifying key elemental components and tracing origin in nuts according to claim 1, characterized in that, In step S3, the data on the content of multiple elements are processed by orthogonal partial least squares discriminant analysis. The top 15 discriminant variables with the highest VIP values ​​are selected as key elements to construct an identification model for the origin traceability of multi-element content analysis, which is used for the rapid identification of the origin of nuts.

10. The method for identifying key elemental components and tracing origin in nuts according to any one of claims 6-9, characterized in that, Step S3 also includes blind sample verification, which includes: selecting multiple nut verification samples determined from different regions, processing the nut verification samples by microwave digestion and obtaining elemental content data by inductively coupled plasma atomic emission spectrometry and inductively coupled plasma mass spectrometry, importing the obtained elemental content data into the multi-element content analysis origin traceability identification model in step S3, and verifying the origin traceability accuracy of the nut verification samples.

Images